The Graph (@graphprotocol): Crypto's Toughest Data Problems (And How The Graph Is Solving Them)

Crypto's Toughest Data Problems (And How The Graph Is Solving Them) Blockchain data is broken in ways most people don't see until they're deep inside the problem. At a recent panel on the hardest challenges in crypto data at Blockworks Digital Asset Summit, a question came up early: Is indexing a solved problem? The short answer is no. The longer answer explains why The Graph has spent seven years building, and why the 2026 Technical Roadmap looks the way it does. Here are the real problems, and where the solutions stand. Problem 1: Standardization is harder than it looks A transaction is a transaction. A transfer goes from A to B. Simple enough on the surface. But across 70+ blockchain networks, different virtual machines, different client implementations, and hundreds of protocol-specific contracts, nothing is really standardized yet. DEX trades alone required integrating against more than 150 protocols before any coherent data format emerged. Every vertical has its own version of this problem. And every time a standard feels settled, a new protocol arrives that doesn't fit. The Graph has been at the center of this since 2018, when Subgraphs became the first widely adopted open standard for indexing smart contract data. The open-source framework that came from that work is now used across web3 by thousands of developers. This reflects a deliberate commitment to open standards as infrastructure that belongs to the public, not to any single provider. The work isn't finished. As institutions enter the space, new standards are needed at the semantic layer: how token assets are classified, how metadata is structured, how data is named and described so that it travels cleanly across organizations and regulatory environments. The Graph 2026 Technical Roadmap directly addresses this, with a continued focus on network-first chain integration, expanded chain coverage, and data services designed to meet institutional requirements out of the box. Problem 2: Speed and scale broke the old model Three years ago, the core user of blockchain data infrastructure was a small to mid-sized DeFi team. Subgraphs served them well. The query volume was manageable. The chains were slow enough that sequential indexing kept up. That model hit a wall. High-throughput chains like Solana, Base, and BSC generate transaction volumes that sequential processing cannot keep pace with. Teams that needed to keep their frontend data fresh, run live analytics, or feed AI agents in real time found that the API-based model had a ceiling. The paradigm shifted from custom APIs to streaming. Substreams was built to solve exactly this. Parallelized processing, cursor-managed reconnection that never drops a data point, and latency in the 600ms to 1900ms range for trading and AI applications. What previously took weeks to reprocess can now be completed in hours. One analytics platform, Amberdata, saw a 72,000% performance improvement over its previous RPC-based infrastructure. The 2026 roadmap advances Substreams toward full protocol integration via Horizon, with trust-minimized payments, a provider selection mechanism, and a decentralized service architecture built for the teams that need blockchain data at the highest throughput. Problem 3: Quality is not the same as consensus Early approaches to blockchain data quality relied on determinism. If multiple independent providers index the same data and return the same result, the output can be trusted. That logic holds in principle. In practice, it has limits. Different Ethereum client implementations return different results for the same transactions. RPC providers have overlapping data sets without the redundancy and diversity that genuine decentralization requires. The underlying assumption that consensus equals accuracy turns out to be more fragile than it appears, especially when the raw input data itself may be inconsistent across providers. The thinking has evolved. The new frontier is cryptographic verifiability. Amp, developed by The Graph core dev Edge & Node, is the product built to clear that bar. A blockchain-native SQL platform with cryptographic verification from the block level up, built-in data lineage, audit-ready provenance, and enterprise-grade deployment options. The goal is a system where data quality is not assumed through agreement but proved through cryptography, and where a compliance team can trace any output back to its source without a blockchain engineer in the room. This is the infrastructure that makes blockchain data usable in regulated environments: payments, treasury oversight, risk management, and AI-driven automation where garbage in means serious consequences on the other side. Problem 4: The customer changed The institutionalization of the industry didn't just change who was buying data. It changed what the data needed to do entirely. The DTCC processes quadrillions in assets annually. It is now confronting a world where it needs position data across hundreds of chains within 5 seconds of each transaction settling. That is not a DeFi startup problem. That is not a problem Subgraphs alone were designed to solve. It is the problem The Graph's 2026 product suite was designed to scale toward. The DTCC's Great Collateral Experiment used The Graph's technology. Major financial institutions are not waiting for the infrastructure to mature. They are building now, and the data layer has to be ready. The Architecture Behind It All Each of these problems requires a different solution. No single indexing approach serves all of them. That is why Horizon matters. Launched in December 2025, Horizon transformed The Graph from a Subgraph-centric protocol into a modular platform where multiple purpose-built data services operate within a single economic and security framework. Subgraphs for structured, queryable APIs across 70+ networks. Substreams for high-throughput real-time streaming. Token API for pre-indexed token data without custom infrastructure. Tycho for real-time onchain liquidity data across DEXs. Amp for verifiable, SQL-native, enterprise-grade data with full audit lineage. Each service purpose-built. All of them operating within one protocol. The hardest problems in blockchain data don't have one answer. They have a stack. And building that stack, openly and at scale, is what The Graph has been doing since the start.